Conquering Customer Experiences with Sentiment & Emotion Analytics | Clarsentia Blog
Conquering Customer Experiences with Sentiment & Emotion Analytics

Conquering Customer Experiences with Sentiment & Emotion Analytics

Posted by Ivan Smith in Marketing & Communications

Your customers are out there talking about your business every day. How do you know that the messages in your public relations, communications and marketing campaigns are resonating and eliciting the response you want? After an incident, how do you know, without guessing, exactly how your customers react?

In my previous article, Top 5 Tips for Building a Sentiment Application, I talked a little about some of the technology and techniques Clarsentia Analytics Solutions uses to analyze and extract sentiment and emotional responses from text. Today's blog provides an example of how you can apply those insights to better understand and serve customer needs.

About Sentiment & Emotion Analytics

We tend to think of purchasing decisions made by customers as decisions firmly cemented in logic: I need a specific product to perform a certain task, therefore I will always buy the one that offers me the best features or the most convenience. While this seems perfectly reasonable on the surface research, has proven time and again that this simply isn’t the case. Most decisions and product selections made by customers are based on how they feel, not what they know -- and it’s understanding those feelings and how they support or undermine our business that we need to understand to build remarkable positive experiences for our customers.

Purchasing decisions made on emotion alone seem counterintuitive. Why would anyone buy something that isn’t the absolute best product or solution for their business? The reason is simple. People, who are the decision makers behind all business, do not simply want a ‘thing’ or a product that solves a problem, they want a relationship with that product or brand. The success of your venture hinges on your ability build those relationships, so you need to understand as much about your target market and customer as you can. That’s why we built Clarsentia, to give businesses the information they need to help build great customer relationships.

The reality is that a lot more is being said about your business than you might think – and there are statistics to prove it. With social data representing 90% of informal feedback on the web and most customers making purchasing decisions based on emotion and likability, you can’t afford to miss not only what your customers are saying but also how they are feeling about your business. Understanding customer reviews alone represents a very small piece of a much bigger picture and you need a holistic approach to capture it. But how can you interpret all the information and make well informed, strategic, data-driven decisions? That’s where Clarsentia comes in.

Traditional social media monitoring services are typically narrow in scope, capturing social data from social accounts your business owns, but how do you know you’ve captured it all? What about those reviews on Google plus or Yelp and that newspaper article that was released by Forbes that captured a turning point for your business? Or what about that tweet where your company was mentioned by a competitor or shareholder but it wasn’t captured on your social media page? Clarsentia continuously adds to its collection of over 900 billion articles from over 20,000 sources stored in a data warehouse.

This data warehouse represents individual newspaper articles, tweets and social media posts about 33,000 businesses – and we add millions of articles each day. The information has been pre-analyzed to understand how the author feels about businesses, which are then aggregated and tracked over time to provide an overall view of an entire industry segment. We give our customers complete access to this data, which they then use to either analyze how their customers feel about the business, or to compare themselves against competitors to better understand how brand reputation improves or degrades over time.

How to Use Sentiment & Emotion Analytics to Improve Customer Experience

So, what do emotional metrics really tell you and how do you quantify emotions in a meaningful way that gives you an understanding of your target market and customers?  To understand better, let’s take a look at some metrics generated on Intel using Clarsentia’s analytics dashboard shortly after the tech giant announced on January 1, 2018 that the Melt Down security flaw existed with its chipset:


As you would expect with an announcement of this nature, the number of publications referencing Intel increased significantly over a short duration. What’s interesting is the increase in emotional intensity expressed, which were mostly disgust, anticipation or surprise. The Clarsentia Dashboard provides added granularity (if we want to know specifically what was said) by clicking any of the above emotional metrics to show all the articles used in our analysis.

Let’s take a look at an article in the “disgust” category published by ABC news on the Intel security flaw:

Above, we see an article was published by referencing Intel that had a high disgust rating (from 0 to 1 where 1 is the highest possible value). By clicking to view the full analysis, we can see exactly which phrases where used that carried “disgust” related to Intel (emotions are highlighted on the emotion wheel to make it easy for our customers to interpret):

We discover that most of the disgust expressed resulted from share sales at Intel.

With this information, we now know with certainty that the audience were very surprised, anxious, and felt disgust when the announcement was made. These feelings gradually decreased as the tech company took measures to respond (including releasing assurances to their customers).

When you’re able to understand the emotions experienced by your customers at a granular level as demonstrated above, it becomes much easier to address their underlying concerns. If we take a look at the graph below which demonstrates the change in public sentiment (positive, negative or neutral) we see that the sentiment change is corroborated by the data:


Here, we get a picture of how a quick decline in sentiment for Intel slides towards the negative in the days leading up to the announcement, with public sentiment hitting a low point about three days after the initial announcement. The Intel example illustrates an important difference between sentiment score and emotional score: while sentiment gives you the black or white (positive or negative) metric, emotional measures look at all the grey areas in between and provide specific psychographics to use for your marketing and customer engagement toolkit.

Why Not Just Survey Your Customers or Read the Articles Yourself?

Companies traditionally try to gather this information through polling or survey techniques which are typically slow, expensive and only offer a single “point-in-time” view of where customer sentiment lies at any given moment. Clarsentia offers a much more holistic view of your target market because we continually add new information as it becomes available, analyze it for you and present it in an easily consumable way. With traditional methods of intelligence gathering, by the time the information is collected businesses are too late to act quickly enough to protect their brand reputation and respond to their customers’ needs. Clarsentia provides instant intelligence gathering so you can respond to your customers quickly.

You might ask why you can’t simply read the information yourself to come up with the same assessment? The answer becomes obvious when you consider how much information you would need to read in order to run a comparative analysis. To give you an idea, here are some stats on what Clarsentia is able to do in a single hour.

Articles Read & Categorized (including tweets, news releases and reviews): 25,000

Sentiment and Emotion Analyses Performed: 334,000 individual sentences

Companies Analyzed and Assessed: 33,000

Clarsentia also leverages proprietary state-of-the-art algorithms to predict where sentiment and emotion measures are heading using machine learning, allowing you to not only see what is going on today, but also the likeliest scenario for  customer satisfaction  in the future. We do not use any third party services, all of our technology is built in-house and proprietary from end-to-end, so you data is totally secure.

You no longer need to read everything out there to understand how your customers feel. The Clarsentia Dashboard provides the overall, global view of the general profile of your customer for you to best adjust your business strategy. Simply log in and view the prepared analyses and instantly know where your market is headed. What could you do with that kind of global view?

The Future of Customer Experience is Sentiment & Emotion Analytics

At Clarsentia we’re able to give our customers immediate, meaningful market insights that can be used to adjust marketing, public relations or communications strategies on the fly much quicker and cheaper than could be done by any other method. Take a look at how Samsung used sentiment analytics to get ahead of competing cellphone manufacturer Apple.

Need more proof? Take a look at how Royal Bank of Canada uncovered a direct link between social sentiment and Chipotles share price. Samsung and RBC understand that emotion metrics are the cornerstone of good customer service and shareholder value, representing a significant shift in the future of marketing and communications strategies.

We hope you enjoyed this article and found the information useful. Contact us for  a no-obligation, free demo of the Clarsentia Dashboard  and let us show you how sentiment analytics can boost your customer satisfaction and engagement.

Enjoyed this Article? Please Share!